T.R | Title | User | Personal Name | Date | Lines |
---|
421.1 | questions | CLOSET::ANKLAM | | Wed May 27 1987 09:30 | 4 |
|
Is this the first job you have run using DOCUMENT T1.0 ? What
Version of VMS are you running, and on what CPU?
|
421.2 | virtual memory questions | VAXUUM::KOHLBRENNER | | Wed May 27 1987 10:17 | 2 |
| And what is your PGFLQUOTA and the system's VIRTUALPAGECNT?
|
421.4 | The data | IJSAPL::KLERK | Theo de Klerk | Wed May 27 1987 13:03 | 10 |
| We are running VAX/VMS V4.5 on a 8650.
It is the first large job that we took from a 750 running BL7 in Munich.
We've installed BL8 and it doesn't any more (haven't tried BL7 as it
was taken off the system here when BL8 was installed).
It did run smoothly on that 750 and BL7.
Our VIRTUALPAGCNT = 70,000 and the process PAGFILQUOTA = 10,000
Theo
|
421.5 | May need more PGFLQUOTA | VAXUUM::KOHLBRENNER | | Wed May 27 1987 13:47 | 4 |
| Try increasing the PGFLQUOTA for your process, to 15000.
bill
|
421.6 | Problem solved, quirk noticed | IJSAPL::KLERK | Theo de Klerk | Wed May 27 1987 14:17 | 13 |
|
The "internal memory" problems sounded to me as if TeX was configured
too small (at least for me). Using your questions on quota, I tried
PAGFILQUOTA of 30,000 and that did the trick. Perhaps 15,000 would
have been good too.
During the processing, I noticed that quite often a <CR> was given between
TeX messages, sometimes leaving me with a blank screen, interchanged with
blocks of "line too long" and "line too short" (it's difficult to please
TeX always!). Is this "blank hopping" usual or will it be suppressed in
the final release?
Theo
|
421.7 | Strange <CR>'s fixed for V10 | WRONGO::PARMENTER | | Wed May 27 1987 14:55 | 9 |
| I don't understand how the configuration of TeX fits into this note
at all, but I must note, that as I write this, I am unable to read note
421.3 which might explain my confusion.
About those errant carriage returns: Yes that is fixed for V10,
but not (alas) for BL08.
David Parmenter
|
421.8 | PGFLQUOTA = 15000 seems adequate | VAXUUM::KOHLBRENNER | | Wed May 27 1987 15:58 | 15 |
| The problem with PGFLQUOTA was occurring in the
tag translator, not in TeX.
For others who are wondering what value is correct for
PGFLQUOTA, be assured that we are wondering the same thing.
We are making some changes to insure that virtual memory
usage stays as low as it can be, and we will be doing some
experiments in the next week to determine what the recommended
quotas should be.
At the moment PGFLQUOTA of 15000 seems to be adequate.
bill
|